Journal of Marine Research, 45, 427-460, 1987
Stoichiometry of C, N, P, and Si fluxes
in a temperate-climate embayment
by S. V. Smithl, W. J. Wiebe2, J. T. Hollibaugh3, S. J. Dollarl, S. W. Hager4,
B. E. Cole4, G. W. Tribble', P. A. Wheeler5
ABSTRACT
Dissolved C, N, P, and Si budgets for Tomales Bay, California, have been used to solve
simultaneous stoichiometric equations which describe a plausible material balance for net
organic matter reactions in the bay. Dissolved Si and P were both exported hydrographically.
Dissolved C and fixed N were imported hydrographically. Ifwe assume that C, N, P, and Si were
supplied to the bay as organic detritus and remineralized at a rate required to balance dissolved
Si and P exports, we can calculate reasonable rates of denitrification and CO2 gas evasion across
the air-water interface. The system is thus interpreted to have been net heterotrophic at the time
of our investigation.
Fluxes attributed to individual components in the system (benthic respiration, water-column
material turnover, biochemical transformations between fixed and gaseous N) were of sufficient
magnitude to account for the system-wide net fluxes, although too noisy to allow piecewise
derivation of net system fluxes.
Denitrification and limitation of primary production by dissolved fixed N in aquatic
ecosystems may be symptoms of other system-scale constraints on net C metabolism, rather than
themselves being system-level controls of net C metabolism.
1. Introduction
Conservation of mass is a fundamental characteristic of nature. Therefore, attention
to the net fluxes of biogeochemically active materials to or from biological systems at
any scale can define constraints on the internal function of those systems. We have
imposed constraints on net fluxes based on independent budgets and inferred chemical
stoichiometry of C, N, P, and Si in a temperate-climate
marine embayment (Tomales
Bay, California; Fig. 1), in order to calculate plausible internal net biogeochemical
reactions.
The study was originally designed to test the hypothesis posed by Smith (1984): The
I. Hawaii Institute of Marine Biology and Department of Oceanography, University of Hawaii,
Honolulu, Hawaii, 96822, U.S.A.
2. Department of Microbiology, University of Georgia, Athens, Georgia, 30602, U.S.A.
3. Tiburon Center for Environmental Studies, Tiburon, California, 94920, U.S.A.
4. U.S. Geological Survey, Menlo Park, California, 94025, U.S.A.
5. College of Oceanography, Oregon State University, Corvallis, Oregon, 97331, U.S.A.
427
428
Journal of Marine Research
[45,2
A
B
0
INNER BAY
OUTER BAY
South
Central
6
o
2
6
8
10
12
14
DISTANCE FROM BAY MOUTH
16
18
(kml
Figure 1. (a) Index map of Tomales Bay, California, showing locations of primary transect
sampling stations. The station numbers denote distance, in km, from the bay mouth. (b)
Longitudinal cross section through Tomales Bay, showing mean depth along the axis of the
bay.
net organic production of marine ecosystems with restricted water exchange is limited
by the availability of P, rather than N, as is more commonly supposed for coastal
marine ecosystems in general (e.g., Ryther and Dunstan, 197 I). It is important to
emphasize that the hypothesis deals with net organic production of the entire system,
not net primary production uncorrected for heterotrophic respiration. The hypothesis
was based on studies in tropical and subtropical oligotrophic systems; Tomales Bay was
chosen as a site to extend that hypothesis into temperate coastal ecosystems.
Tomales Bay receives more dissolved and particulate terrigenous input than the
sites previously examined and also derives materials from coastal upwelling. We will
Smith et al.: Stoichiometry of C, N, P
1987]
&
Si fluxes
429
Table 1. Dimensions ofthe major sectors of Tomales Bay.
Length
m
Width
m
Depth
m
Area
106 m2
Volume
106 m3
6000
1700
2.2
10.2
22.4
Inner
Central
South
10000
4000
1400
1000
4.4
1.0
14.0
4.0
61.6
4.0
Total
20000
1410
3.1
28.2
88.0
Sector
Outer
demonstrate that these inputs rendered the intended hypothesis untestable in Tomales
Bay, at least at the time of our study. Nevertheless, we derive important conclusions
about ecosystem-scale nutrient dynamics and biochemical pathways in the bay. We do
so by addressing the following questions.
Was this embayment a net producer or net consumer of organic carbon at the time of
our study? What were the relationships of N, P, and Si fluxes to C metabolism?
Finally, we have compared inferences drawn from the system-scale budgets with
smaller-scale experiments designed to evaluate benthic fluxes, water-column nutrient
turnover, and biochemical transformations between fixed and gaseous N. The purpose
of these experiments was to verify that reactions along the inferred biogeochemical
pathways were of sufficient magnitude to account for net baywide fluxes.
We have attempted to be comprehensive in establishing the hydrographic and
biogeochemical fluxes of nongaseous dissolved forms of each element analyzed (C, N,
P, and Si). The budgets also deal implicitly with gas flux (N2, N20, CO2) across the
air-water interface and with particle flux, as consequences of stoichiometry inferred
from the dissolved-load budgets. We emphasize that our investigation does not resolve
the significance of annual or longer-term variation in the mass balances of biogeochemically active materials.
2. Study site-Tomales Bay
Tomales Bay (Fig. 1) is well suited for mass balance studies of a relatively
unperturbed system. The watershed (an area of about 600 km2) has a population of
about 3000 persons; human pressures, in the form of nutrient pollution, other
environmental degradation, or fishing pressure, are relatively minor. Runoff from
pasture land may significantly affect the wet-season nutrient budgets, but was minimal
during our study.
The geological setting and physiography of Tomales Bay have been described by
Daetwyler (1966). The bay, which occupies a rift valley formed by the San Andreas
Fault, approximates a long, narrow rectangle (20 km by 1.4 km). The average depth is
3.1 m (zma. = 19 m). About 30% of the area is shallower than 2 m, with shoals near the
mouth, the head, and the northeast shoreline. Between 4 and 16 km from the mouth is a
silled basin.
Figure 2. Summary of long-term mean stream flow (Lagunitas and Walker Creeks), evaporation, and rainfall in the Tomales Bay watershed; data summarized from various sources.
The shoals near the mouth support extensive eelgrass (Zostera marina) beds;
limited eelgrass and benthic algae persist along the northeastern shore, approximately
16 km into the bay. Benthic plant populations are not conspicuous below about 2 m
(except for some giant kelp, Macrocystis pyrifera. in the Outer Bay). Benthic diatoms
may be present on much of the bay floor, especially on the shoals near the southern end
of the bay. However, based on suspended chlorophyll levels, we believe that
phytoplankton dominate the submerged plant biomass of the Inner Bay. The bay floor
supports abundant infauna. Macrozooplankton were abundant at the mouth of the bay;
except for localized swarms of the jellyfish, Aurelia. macrozooplankton were rare
within the Inner Bay during the present study.
E. H. Smith et al. (1971) discussed aspects of water composition and circulation.
Exchange with the ocean is predominantly tidal, with wind-induced mixing in the
Inner Bay. The mean semidiurnal tidal range is 1.1 m.
Two major streams enter the bay (Fig. 1). Walker Creek flows into the north end of
1987]
Smith et al.: Stoichiometry
431
of C. N, P & Si fluxes
the bay along the eastern shore; its effects on salinity are minor, episodic, and localized
(Smith et al., 197 I). Lagunitas Creek, which flows into the southern end of the bay,
can substantially depress surface salinity of the Inner Bay during the rainy season.
During the dry season, bay salinity is near oceanic values, with insignificant vertical
salinity stratification. Information on long-term rainfall, evaporation, and stream flow
is summarized in Figure 2. Major stream flow occurs from about November to April,
with little runoff and significant net evaporation during the remainder of the year.
Coastal upwelling is strong in the vicinity of the mouth of the bay, bringing
nutrient-rich water to the sea surface. The upwelling is episodic, tied to strength and
persistence of northerly winds, but is strongest between April and September (Bakun,
1975; Abbott and Zion, 1985).
3. Experimental design
We have used a steady-state, one-dimensional box model to calculate material mass
balances in Tomales Bay. The model is intermediate between the river-dominated
model employed by Boyle et al. (1974), Wollast (1978), Officer (1979), Kaul and
Froelich (1984), and others, and the meteorologically dominated model of Smith and
10kiel (1978) and Smith and Atkinson (1983).
Consider a system of constant volume (V), open to exchange of ocean water by
advection and mixing, with freshwater advection (F) from land, and with net
evaporation minus rainfall (E) across the system. Let dS/dt = salinity change with
time; So ~ oceanic salinity; SJ = stream salinity; S. = salinity of rain water and water
removed by evaporation; dS/dX = longitudinal salinity gradient (positive into bay);
Ax = cross sectional area of the system perpendicular to X; and Kx = longitudinal
mixing coefficient. E, Kx, Ax, and S are all functions of X:
V· dS/dt
= (E - F) • So
=
+ E • S. + F·
SJ - Kx . Ax • dS/dX
(1)
0 (at steady state).
Ifwe assume that streams and rainfall minus evaporation both involve water flux of
salinity = 0, the equation simplifies:
V. dS/dt
= (E - F) . So - Kx . Ax . dS/dX
= O.
(2)
If all other terms in Eq. 6 are known:
(3)
A general term (By) can be added for any material (Y) which has fluxes that are not
conservative with respect to salinity; both By and Yare functions of X . Other notation
is analogous to that presented for salinity. The following equation can be derived
(assume Y. = 0, but make no such assumption for YJ):
V· dY/dt
=
(E - F) • Yo
+ F·
YJ - Kx . Ax . dY/dX
+ By ~
O.
(4)
Journal of Marine Research
432
[45,2
It follows from Eqs. (3) and (4):
(E - F) . (Yo - So . dYjdS)
+ F· YJ + By
This equation can be broken into non conservative processes
Y in ocean water «E - F) . Yo) and river water (F·
(-(E
=
O.
(5)
(By), advection of
YJ), and mixing
- F) . So· dYjdS).
For evaluating By, Eq. 5 is scaled by map area (Am):
(6)
Eq. 6 describes ~ in an embayment between the oceanic salinity isopleth and the
bay head. E and F are obtained landward of that isopleth. Values for ~ can be derived
landward of any isopleth, and the solution within any bay sector follows by integration
between isopleths. The appropriate gradient, dY j dS, is the gradient in the vicinity of
the isopleth which bounds the flux between the embayment and ocean (Officer,
1979).
Calculation of By is sensitive to the form of the equation used to describe Yas a
function of S (Boyle et al., 1974). This limitation has its most severe effect in
examining variation of By along the bay; we restrict calculations to bay-wide fluxes,
rather than sector fluxes.
Net hydrographic exchange time in the system (T) can be calculated from the
system volume (V), daily water flux (E - F), average bay salinity (8;,), and incoming
oceanic salinity (So) (modified from Smith and Jokiel, 1978; Pilson, 1985):
(7)
Our sampling period (29 September-9 October 1985) was chosen to be near the end
of the dry season in order to minimize advection and maximize the effect of By on
horizontal concentration gradients (Eq. 6). We recognize that such a "snapshot in
time" does not characterize temporal trends in a seasonal system (e.g., Kaul and
Froelich, 1984), and we plan to extend the study to an annual cycle.
4. Materials and methods
This investigation consisted of five coordinated efforts: (1) hydrographic sampling
for budgetary analyses; (2) field studies of benthic nutrient flux: (3) water-column
primary productivity measurements; (4) water-column Nand P turnover experiments;
(5) nitrogen fixation and denitrification measurements.
a. Hydrographic studies. Figure 1 shows the primary stations for transect sampling
(numbered by distance, in km, from the bay mouth). Primary sampling, undertaken 5
times over a 10-day period, included salinity, temperature, dissolved inorganic and
organic nutrients, pH, and total alkalinity (T A). The normal sampling suite included
1987]
Smith et al.: Stoichiometry ofe, N. P & Sifluxes
433
stations 0 through 18, in the navigable portion ofthe bay. Station 20 was sampled only
once. We also sampled once for dissolved organic carbon (DOC) and chlorophyll a.
Samples were also collected for establishing the C:N:P ratio of particulate matter in
the water column, major submerged benthic plants in the system, and sediments. In
addition to the bay samples, we once collected water from the two streams flowing into
the bay.
Salinity and temperature profiles (obtained with an Applied Microsystems STD)
established the lack of vertical stratification. Subsequently, surface temperatures were
taken with a bucket thermometer; surface samples were taken for measurement with a
Plessey model 6230N laboratory salinometer standardized against a secondary standard (±O.OI0j0o). Samples for all analyses were collected by subsampling from a
bucket.
pH was measured with a Cole Parmer model 5985-00 pH meter and Orion model
8166BN Ross combination electrode, within two hours of sample collection. pH of
water samples held in air-tight containers near collection temperature showed negligible drift over this period. Alkalinity (TA) was measured with the same system,
generally following the technique of Smith and Kinsey (1978). Precision within
analysis sessions was better than 0.006 pH units for both pH and TA. pH, not explicitly
reported here, varied between 8.0 and 8.3. pH and TA were used to calculate dissolved
inorganic C (DIC) and CO2 partial pressure. Lyman's dissociation constants (in
Skirrow, 1975) and Weiss's (1974) CO2 solubility data were used.
Samples for DOC were filtered into Teflon-capped quartz test tubes and frozen until
analysis. After persulfate oxidation, samples were analyzed on an Oceanography
International model 700 TOC analyzer, according to procedures in the instrument
manual.
Dissolved N, P, and Si analyses were conducted by the U.S. Geological Survey,
Menlo Park (USGS), and the Hawaii Institute of Marine Biology Analytical Services
group (UH). Automated analysis procedures by the two laboratories were similar,
generally following the Technicon procedures for N03-, N02-, NH4+, PO/-, and Si,
as locally modified by those laboratories (details available from SVS [UH] and SWH
[USGS]). Both laboratories use ultraviolet oxidation (Armstrong et al., 1966) for total
dissolved Nand P analyses. Dissolved organic Nand P (DON, DOP) are calculated by
difference between total and inorganic (DIN, DIP) nutrient concentrations.
Pre-analysis sample handling differed between laboratories. The 30 September
samples and a set of samples collected on 3 October for intercalibration were analyzed
by the USGS. Other nutrient analyses reported were analyzed by UH only. USGS
samples were held on ice in 2-liter Nalgene containers for 3-6 hours before filtration.
After filtration through 0.45 #Lm Nuclepore filters, samples were refrigerated until
analysis within one day of collection. UH samples were immediately filtered through
glass fiber filters, into acid-washed, sample-rinsed polyethylene bottles. Samples were
held on ice for up to an hour, then frozen for 2-3 weeks until analysis. Agreement
434
Journal of Marine Research
[45,2
between laboratories ranged from about 2% (Si and PO/-), to 10% (N03- + N02-,
NH/). The USGS data for 30 September were adjusted by regression analysis for
intercomparison with the UH data.
Particulate C and N analyses were conducted with a Hewlett Packard modelI85B
CHN analyzer. One aliquot of each sample was held at 480-500°C for 4 hours in a
muffle furnace to remove organic C, for subsequent analysis of CaC03-C; this value
was subtracted from the total C measured on an unmuffled aliquot, to yield organic C.
A third aliquot was analyzed for total P after ashing (as above) and digestion of the ash
in 6N HCl.
b. Benthic flux studies. Benthic nutrient fluxes were measured at two sites within each
bay sector. Measurements were made in cube-shaped opaque fiberglass chambers
30 cm on a side, and open on the bottom. A hard rubber flange rimming the chamber
bottoms penetrated 8 cm into the sediment. Lead collars anchored the chambers. Six
replicate incubations were conducted at each site. Battery-powered stirrers in each
chamber prevented water stagnation.
Chambers were lowered to the bay floor from a boat; divers positioned the chambers
in smooth areas approximately 2-3 m deep and devoid of sea grass. Divers extracted
water from each chamber into plastic syringes via a stopcock, temporarily opening a
second stopcock to replace the sample water volume with ambient bay water. Dark
bottles with bottom water served as water-column controls. Samples were extracted a
few minutes after chamber deployment, then approximately 8 and 24 hours later.
Samples were collected for pH, TA, and dissolved nutrients; analytical procedures
followed those used for transect studies. Because of sediment resuspension during
chamber emplacement, fluxes were calculated with the 8 and 24 hour data.
c. Primary productivity. Samples for productivity analyses were collected from
surface waters at one site in each of the bay sectors. After inoculation with 5 #LCi of
NaH14C03 in each 150 ml bottle, 24 hr incubations were undertaken in incubation
chambers floating nearshore. Samples were exposed to 8 intensities of ambient
irradiance (100, 55, 30, 15,8,3, 1,0%), by placing bottles in tubes of different mesh
screen. Following incubation, 3 ml aliquots from each bottle were placed in scintillation vials. To remove residual14C, 0.1 ml of 0.2N HCI was added to each vial, which
was shaken uncapped for 30-60 minutes. The activity of each sample was measured by
liquid scintillation. Data were corrected for dark-bottle uptake.
Daily productivity through the water column was calculated as discussed by Cole
and Cloern (1984), fitting the light versus productivity data to a hyperbolic tangent
function. 24-hour chlorophyll-specific respiration rate was assumed to be 10% of the
maximum carbon assimilation rate. Net primary productivity was calculated as the
1987]
Smith et al.: Stoichiometry
of C, N, P & Si fluxes
435
sum of water-column productivity minus integral respiration over the mean water
depth for the corresponding sector of the bay (Table 1).
d. Water-column
Nand
P cycling. Samples for P and N regeneration experiments
were collected by immersing clean 20 liter plastic carboys to a depth of 10 em. The
carboys were returned to the laboratory and held in dim light (100 ~E m-2s-1 PAR
from Cool-white fluorescent tubes) until experiments were run later on the collection
day. To initiate experiments, water was mixed by swirling the carboy, and subsamples
were poured through a 100 ~m mesh screen into 1 liter or 10 liter plastic bottles.
The I liter bottle was spiked with carrier-free 32p_PO/- (approx. 10 ~Ci; exact spike
determined by radioassaying a 1.0 ml subsample). P regeneration rates were determined by following the time course of PO/- specific activity. At various times during a
four-hour time course, 100 ml samples were filtered and frozen for PO/- analysis.
Filters were rinsed three times with 10 ml portions of unlabelled sample, placed in
vials, and radioassayed by liquid scintillation.
Specific activity of PO/- was determined by extracting the phospho-molybdate
complex formed in the standard PO/- analysis with isobutanol (Strickland and
Parsons, 1972). 32pin the extract was determined by assaying a 1 ml portion. Extract
absorbance was measured by spectrophotometer, in order to determine concentration.
P uptake and regeneration were measured in one subsample per carboy, and one
sample was filtered at each time point. Specific activity determinations at each time
point were made in triplicate. Uptake and regeneration were calculated from linear
(type 1) regressions, using log-transformed specific activity data. Calculations
followed Harrison (1983).
Incubations for NH4 + uptake and regeneration were carried out in the 10 liter
carboy, spiked with enough 15N-NH4+ to raise NH4 + concentration by approximately
5%. Subsamples were withdrawn after tracer addition and after 1,2, and 4 hours.
The relative abundance of 15N in NH4 + was determined on filtrates (GF IF
Whatman filters) by conversion of NH4 + -N to indophenol-N and subsequent extraction of the indophenol into dichloromethane. After evaporation of the solvent, the atom
percent 15Nof the samples was determined by emission spectrometry (after Dudek et
al., 1986). Rates of NH4 + regeneration were calculated according to Blackburn (1979)
if there was a significant change in NH4 + during the incubation, and Glibert et al.
(1982) if there was not. Rates of NH4 + uptake were calculated as the sum of the
change in concentration and the rate of regeneration (Blackburn, 1979).
e. Nitrogenfixation-denitrification
studies. Preliminary screening incubations were
conducted for N2 fixation and denitrification. Samples for screening included
sediments, rooted Zostera, and drift grass or algae. Incubations on undisturbed
436
Journal of Marine Research
35.0
,.-..
g
.....•..
•
•
c
0
34.5
a
.•....•.•
~
z
~
~
34.0
33.5
(/)
33.0
o
30
03
07
08
09
SEPT.
OCT.
OCT.
OCT.
OCT.
••••
2
4
[45,2
.~
i!
0
i
~
lJ.
.
6 8 10 12 14 16 18 20
DISTANCE FROM BAY MOUTH (KM)
Figure 3. Salinity as a function of distance from the mouth of Tomales Bay. SeptemberOctober 1985.
sediment cores were conducted in small glass cylinders stoppered at both ends; grass or
algae incubations were conducted in larger cylinders.
Nitrogen fixation measurements generally followed Webb et al. (1975). Ten
replicates were run for each site. Acetylene reduction was measured on an A.I.D.
Instruments gas chromatograph equipped with a 6-foot column packed with Poropak
R and a flame ionization detector. Only relative rates of N2 fixation are reported.
Potential denitrification (i.e., the ability of the microbial community to utilize N03present in a concentration of approximately 1 mM) was measured in one of two ways.
First, N03 - loss was measured, with the assumption that all N03 - loss could be
attributed to denitrification. Six samples were run by this method at one site. Second,
N20 was added to samples in the presence of acetylene to block N20 reductase
(Oremland et al., 1984); N20 loss was measured on a Carle gas chromatograph with
an 18 foot Poropak Q column. A time course was run with 8 samples.
5. Results
a. Hydrographic data. Between 0 and 6 km (Outer Bay), there was little gradient in
salinity (Fig. 3) (mean ± std. dev. = 33.44 ± 0.06%0), because Kx is large relative to
E - F (Eq. 3). Over the next 10 km (central sector), salinity rose to 34.7%0,reflecting
long water residence time and net evaporation. Salinity in the south sector reached
34.80/00due to net evaporation (E > F), and then dropped slightly due to runoff from
Lagunitas Creek (E < F). Salinity was constant (±0.05°/00) with depth; samples
transverse to the longitudinal profile showed only minor variation. The salt-balance
model is not applicable in the Outer Bay because dS / dX = 0, so we restrict calculations of nonconservative fluxes to that portion of the bay inside the 33.5%0isopleth
(approximately station 6). Volume-weighted average Inner Bay salinity was 34.07%0.
Water temperature averaged 14.5°C in the Outer Bay, 17.0°C in the central sector,
and 18.4°C in the southern sector during our study. Area-weighted Inner Bay average
temperature (used for calculating evaporation rate, Table 2) was 17.2°C.
Smith et aJ.:Stoichiometry ofe, N, P & Sij/uxes
1987]
Table 2. Weather,
October 1985.
water
temperature,
and stream
runoff
into Tomales
437
Bay, September-
A. Mean weather conditions
Pt. Reyes*
ALERTsta.
Variable
Wind (m S-I)
Humidity (%)
Air temp. (0C)
Rain (mm d-')
Inner Bay
Water temp. (°C)t
Evap. (mm d-I)tt
Bodega***
Coast Guard
Marshall**
4
5
80
15
0.1
Mean
3
4
85
16
90
17
17.2
1.6
B. Inner bay water budget
Net evaporation-rainfall:\:
Lagunitas Creek stream flow:j::j:
Net ocean-water replacement
2.88 . 104 m3 d-I
1.30 . 104 m3 d-I
1.58 . 104 m3 d-I
*Remote recorder, National Weather Service. Data from 25 September through 9 October
1985.
**Private 000.,2-3 obs./day; NE side of bay. Data from I September through 9 October.
***6 obs./day. Data available from National Weather Service.
tData from 30 Sept.; 3, 7, 8,9 Oct.
ttFormula
in Smith and Jokiel (1978), Sverdrup et al. (1942).
:\:Calculated for area of 18 . 106 m2 (Table 1).
:j::j:Data from U.S. Geological Survey.
Table 3. Cubic regression equations for nutrients as functions of salinity in Tomales Bay. The
equations are of the form Y = a + b . 8 + c . 82 + d . 83, where Yis the nutrient and 8 is
salinity. Values and slopes of Yat 8 = 33.5 0Ax>
and the concentration of Yin Lagunitas Creek
water (Yf) (October 1985) are also given for modelling purposes.
Material
(Y)
Si
PO/DOP
N03N02NH4+
DON
TA
DIC
DOC
a
-31693.70
-821.23
-76.73
3125.64
171.20
737.00
-1055.41
38638
32113
838
b
835.84
c
d
r2
Y".•
dY/d8
JtM
JtM/°Ax>
l'f
JtM
32.2495
-0.863820
0.98
23.30
88.29
20.740
1.942
0.8475
0.0805
-0.021886
-0.002082
0.99
0.96
1.85
0.39
3.84
0.33
0.5
0.3
-79.759
-4.527
-23.049
27.192
-3.2917
-0.1723
-0.4811
1.0313
0.086303
0.004631
0.015399
-0.026727
0.95
0.98
0.78
0.99
4.20
0.29
3.88
8.09
-9.74
-0.48
-3.44
6.31
1.1
0.0
0.7
8.1
1.0228
0.7509
0.0485
0.97
0.85
0.93
-1094.80
-997.40
-20.06
-33.914
-22.085
- 1.726
2355
2146
52
76
51
28
190
1980
2010
260
438
Journal of Marine Research
[45,2
70
r--.
0
(0)
60
I
50
~
:::!:
::t 40
......•.
Vi
30
t:.
•
20
10
II
0
0
2
•i •I
6
t:.
0
•
•
c
0
l!.
4
I t:.8
30
03
07
08
09
•!
SEPT.
OCT.
OCT•
OCT.
OCT.
8 10 12 14 16 18 20
DISTANCE FROM BAY MOUTH (KM)
70
60
(b)
r--. 50
:::!: 40
::t
..•..••.30
Vi
20
10
o
33.4
33.8
34.2
34.6
SALINITY (0/00)
Figure 4. Si in Tomales Bay, September-October 1985. (a) Daily data as a function of distance
from mouth. (b) Mean Si at each station as a function of salinity, with cubic regression line.
Station mean nutrient versus salinity data were fitted with cubic regression
equations. The cubic equations allowed for curvature in some data sets. The regression
coefficients and nutrient concentrations (Yo) and slopes at 33.50j0o (dYjdS) are
presented in Table 3. Values for stream concentrations (Yf) are also tabulated.
There was no spatial Si gradient in the Outer Bay, but there was day-to-day
variation (Fig. 4a) (mean ± std. dev. = 18 ± 7 ~M). This variation apparently
reflected variable delivery of upwelled water to the bay. From 6 km southward, Si
showed a general increase to about 60 ~M (at 12 km), then decreased. The
volume-weighted mean Si concentration in the Inner Bay was 44 ~M. When station
means Si data between 0 and 18 km are presented as a function of salinity, the data are
well approximated by the cubic regression line (Fig. 4b). Station 20 (where stream
runoff lowered salinity slightly) is excluded from the Si-salinity regression analysis, as
it is for other regression analyses.
The PO/- trend (Fig. 5) was similar to Si, although without a prominent decrease at
the southern end of the bay. Outer Bay PO/- was also temporally, but not spatially,
1987]
Smith et al.: Stoichiometry of C. N, P & Si fluxes
439
5
(0)
..---.. 4
I
~
::t
'-" 3
I
1"1"<t 2
~
0
• Ii •
i
a
a...
I •
0
•
0
a
0
•
0
2
Cl
6.
4
6
30
03
07
08
09
SEPT.
OCT.
OCT.
OCT.
OCT.
8 10 12 14 16 18 20
DISTANCE FROM BAY MOUTH (KM)
1.0
,
,
,
(b)
0.8
• • •t:. •~ •B
8
..---..
~
0.6
::t
'-"
a...
•CI •
0.4
a
0
0
0.2
0
~
0
2
4
6
0.0,_
o
Ii
0
~
[]
0
0
8 10 12 14 16 18 20
DISTANCE FROM BAY MOUTH (KM)
..---..
5
~
::t 4
'-"
a... 3
(c)
0
w
~
a
2
(/)
(/)
DOP
-----~-----~--~~
Ci
0
33.4
33.8
34.2
34.6
SALINITY (0/00)
Figure 5. P in Tomales Bay, September-October 1985. (a) Daily POl- as a function of distance
from mouth. (b) As for (a), but DOP. (c) Mean POl- and DOP as functions of salinity, with
cubic regression lines.
variable (1.5 ± 0.5 JtM), again apparently reflecting variation
PO/- increased with increasing salinity, to about 4.5 JtM. Inner
PO/- concentration was 3.3 JtM. DOP increased from 0.4
averaged mean = 0.5 JtM).
Four N species are shown as functions of location (Fig. 6a-d)
in coastal upwelling.
bay volume-weighted
to 0.6 JtM (volumeand salinity (Fig. 6e,
440
Journal of Marine Research
16
,......
::::E
12
6
6
::t
'-'
I
r")
0
z
•
(0)
0
•
c
6
6
8
i
4
0
0
0
•
2
24.0
30 SEPT .
03
07
08
09
OCT.
OCT.
OCT.
OCT.
::::E
~
z
12.0
0
~
iii ~
•
4
8 10 12 14 16 18 20
6
(d)
18.0
,......
0
•
Cl
6.0
0
0.5
•
0.4
€I
::::E
~
I
N
0
z
0.3
iii ~
0.2
0.1
0.0
::::E
e e
•
• a•
~
z
•8
0
6
0
2
4
6
6
iii
=
0
0
D
A
••
~
0
Cl
I.LJ
~
0
Vl
Vl
Ci
,......
::::E
e
::t
'-'
+-.t
4.0
6
~
::t
•
6
Z
•
Cl
I.LJ
6
••
0
2.0
0
6
0
6
III
•
0
0.0
0
2
4
6
•
2
4
6
8 10 12 14 16 18 20
(e)
4
3
2
1
34.2
14
12
'-' 10
6
!0
6
SALINITY (0/00)
::::E
0
I
z
~
e
6
33.8
,......
6.0
II
o •8 ~
~
5
8 10 12 14 16 18 20
(c)
0
6
DISTANCE FROM BAY MOUTH (KM)
8.0
•
DISTANCE FROM BAY MOUTH (KM)
,......
(b)
i0
•~
0.0
DISTANCE FROM BAY MOUTH (KM)
,......
[45,2
~
8
6
0
(f)
~
0
8 10 12 14 16 18 20
DISTANCE FROM BAY MOUTH (KM)
Vl
Ci
2
0
33.4
---------- ---.--T33.8
34.2
34.6
SALINITY (0/00)
Figure 6. N in Tomales Bay, September-October 1985. (a) Daily NO] - as a function of
distance from mouth. (b) As for (a), but N02 -. (c) As for (a), but NH4 +. (d) As for (a), but
DON. (e), (0 mean N species as functions of salinity, with cubic regression lines.
f). NO]-, N02-, and NH4+ were elevated and variable in the Outer Bay (N03- =
5.1 ± 3.7 J.tM; N02 - = 0.34 ± 0.13 J.tM; NH4 - = 3.9 ± 1.1 J.tM). The high and variable
N03 - values also apparently reflect the variable influence of upwelling. Evidence of
rapid cycling of fixed N is seen in the tendency of both NH4 + and N02 - to be elevated
in Outer Bay water samples. All forms of DIN decreased in the central sector, with
some increase in the southern sector (volume-weighted mean: N03 - = 1.6 J.tM;
Smith et al.: Stoichiometry of C. N, P & Si fluxes
1987]
2700
.......
•...•
•
0
2600
I
•c
>
'5r:F 2500
A
,
30 SEPT.
03 OCT.
(0)
~
07 OCT.
08 OCT.
09 OCT.
0
Q)
::t
.•......
2400
~
o ~ ~
2300
~
o
!
2
••
a•
•~
!I
••
i
A
!
4
6
•
"~
0
i
~
2600
,-...
•...•
I
2500
>
'5
r:F
!
.•......
2400
~
·
2300
33.4
8 10 12 14 16 18 20
DISTANCE FROM BAY MOUTH (KM)
33.8
2300
~
::t
.•......
2200
0
o
I
0
~ ~
• i= •
2
6
•
••
2100
2000
o
4
(e)
•
~
~
• 8~
•
110
(c)
,-...
90
::t
.•......
80
•
u
70
c
60
0
•
50
40
0
•
2200
2100
33.4
8 10 12 14 16 18 20
100
•
2300
•
••
••
2 4 6 8 10 12 14 16 18 20
DISTANCE FROM BAY MOUTH (KM)
34.2
34.6
110
•
•
33.8
SALINITY (0/00)
DISTANCE FROM BAY MOUTH (KM)
~
34.6
SALINITY (0/00)
(b)
u
0
34.2
2400
2400
.......
~
441
,-...
90
::t
.•......
80
u
70
c
60
~
0
•
(f)
100
•
50
40
33.4
33.8
34.2
34.6
SALINITY (0/00)
Figure 7. C in Tomales Bay, September-October 1985. (a) Daily TA as a function of distance
from mouth. (b) As for (a), but DIC. (c) As for (a), but DOC. (d), (e), (f) Mean TA, DIC,
DOC as functions of salinity, with cubic regression lines.
N02 - = 0.2 ~M; NH4 + = 2.5 ~M). Salinity-DIN trends are summarized in Table 3.
DON increased from 7 ~M in the Outer Bay to 13 ~M (mean = 10.8 ~M).
When the Si, PO/-, and DIN data are viewed together, the upwelling influence is
even more apparent. Water samples from the Outer Bay show positive correlations
between all combinations of these materials (r > 0.95), with the concentration ratios
being about the expected ratios for variable fractions of upwelled water along the west
442
Journal of Marine Research
[45,2
Table 4. Calculated steady-state net fluxes of dissolved Si, P, N, and C, Inner Tomales Bay, late
summer 1985.Values are in mol d-I (mmol m-2 d-I). See text for detailed explanation.
Advection
Material
Si
River
Ocean
Mixing
N onconservative
+2470.0
+368.1
-46731.9
+43893.8
( +2.439)
PO/DOP
tot. P
+6.5
+3.9
+ 10.4
+29.2
+6.2
+35.4
~2032.5
-174.7
-2207.2
+ 1996.8
+ 164.6
+2161.4
( +0.111)
( +0.009)
( +0.120)
N03N02NH+
DIN
DON
tot. N
+ 14.3
+0.0
+9.1
+23.4
+ 105.3
+ 128.7
+66.4
+4.6
+61.3
+ 132.3
+ 127.8
+260.1
+5155.4
+254.1
+1820.8
+7230.3
-3339.9
+3890.4
-5236.0
-258.6
-1891.2
-7385.8
+3106.8
-4279.2
( -0.291)
( -0.014)
( -0.105)
( -0.410)
( +0.172)
( -0.238)
TA
DIC
DOC
tot. C
+25740
+26130
+3380
+29510
+37209
+33906
+822
+34728
-40227
-26994
-14820
-41814
-22722
-33043
+ 10618
-22425
( -1.262)
( -1.836)
( +0.590)
( -1.246)
coast of North America (Park, 1967; Fehler, 1981). There were negative correlations
between each inorganic nutrient and water temperature in the Outer Bay.
The data for TA, DIC, and DOC are shown in Figure 7. TA increased from 2300 to
2600 ~eq I-I (mean = 2435 ~eq 1-1). DIC increased from 2100 to 2300 ~M
(mean = 2203 ~M). DOC increased from about 40 to 100 ~M with increasing salinity
(mean = 70 ~M). CO2 partial pressure (not illustrated) did not show a clear spatial
trend, but averaged about 470 ~atm throughout the bay.
b. Budgetary calculations. The water budget defining E and F is summarized in
Table 2. Table 4 presents calculated fluxes (from Eqs. 5 and 6) for each nutrient south
of the 33.5%0isopleth.
The water budget was dominated by evaporation (Table 2), resulting in salinity
elevation in the Inner Bay. The margin of E over F was 1.6 . 104 m3d-1 (0.02% of the
Inner Bay volume). Inner Bay hydrographic exchange time during this "diffusive
regime" driven by tidal and wind mixing was about 78 d (Eq. 7). E altered salinity
sufficiently that water exchange could be estimated, but was insufficient to affect the
exchange rate. The cross-sectional area at station 6 is 5800 m2, and the local salinity
gradient was 1.6 • 10-5°/00 m-I• Eq. 3 yields a longitudinal mixing coefficient of
5.7 . 106 m2d-l• If, instead of station 6 statistics, we use Inner Bay average statistics
(dS/dX = 8.1 . 10-5°/00 m-I; Ax = 4700 m2), Kx decreases to 1.4. 106 m2d-J•
During the wet season, F can exceed 106 m3d-1 (Fig. 2), a flow rate which would
1987]
Smith et aJ.:Stoichiometry ofC, N, P
&
Sijiuxes
443
directly affect water exchange. Conservative exchange time would decrease (e.g.
Pilson, 1985), and advection would decrease the system signal from By (Eq. 5).
Variability of Yo in response to upwelling might introduce apparently nonconservative behavior in conservative properties of the system (Loder and Reichard, 1980;
Officer and Lynch, 1981). Inspection of Figures 3-7 suggests that this variability is
largely damped out from station 6 southward for most variables.
The Inner Bay showed hydrographic export of dissolved Si, implying a dissolved Si
source within the bay. We assume that there is no significant aerial Si source or sink.
The net release rate was 2.44 mmol m-2d-1 during our study (Table 4). PO/- was also
exported hydrographically, at a rate of approximately 0.11 mmol m-2d-1 (Table 4).
Again, we assume no significant aerial source or sink. DOP export was an additional
0.01 mmol m-2d-1•
Unlike Si and P, which were being exported from the Inner Bay, DIN was imported
hydrographically (Table 4). Net DIN influx totalled 0.41 mmol m-2d-1, with N03accounting for about 70% and NH4 + most of the remainder. NO] - was taken up before
NH/ (Fig. 6). DON export (0.17 mmol m-2d-1) accounted for about 40% of the DIN
import; there must have been a net dissolved fixed N sink totalling 0.24 mmol m-2d-l•
There can be aerial input of fixed N (via fixation of N2 gas) and a dissolved fixed N
sink other than organic matter (denitrification of NO] - ).
Inner Tomales Bay showed net hydrographic input ofTA and DIC, and net export
of DOC (Table 4). The total C budget includes the additional term of CO2 gas flux
driven by the CO2 partial pressure differential. The observed air-to-sea CO2 partial
pressure differential (340--470 #Latm)indicated CO2 gas loss from the water to the
atmosphere.
c. Stoichiometric implications of the biogeochemical budgets. The system-scale
budgets are now considered simultaneously in order to derive a plausible stoichiometry
of biogeochemical reactions involving C, N, P, and Si. Nixon and Pilson (1984) and
Hammond et al. (1985) have presented similar derivations.
Export of dissolved Si and PO/- suggests that the Inner Bay (water column plus
sediments) was heterotrophic at the time of our study, showing net P release from
particulate organic matter and Si release from diatom frustules. Alternative explanations might be that P was released from inorganic matter by redox-mediated
desorption, and that Si was released by silicate mineral dissolution. However, Si and
PO/- concentrations for all samples in the bay were correlated (r = 0.87; n = 63),
suggesting that the source for Si and P was similar. The high CO2 partial pressure of
the water indicates that CO2 gas escaped to the atmosphere, consistent with regeneration from biogenic material.
Organic matter supporting this nutrient regeneration might have come from land
(during winter runoff), from organic production sedimented earlier in the season, or
from input of organic matter across the Outer Bay sill. A time series of budgets would
be needed to resolve among these alternatives.
444
Journal of Marine Research
[45,2
There are possible complications to the general model we propose. Martens et al.
(1978) and Klump and Martens (1981) observed that there can be differential loss of
dissolved P to mineral phases or to adsorption during oxidation of organic matter.
Krom and Berner (1981) argued that much of the P released from sediments is derived
from inorganic material. However, Watanabe and Tsunogai (1984) pointed out that,
while the immediate source of P released from the sediments may be desorbed
inorganic P in response to varying redox conditions, the P is likely to have originated
from organic material deposited in the sediments and oxidized there. Sediment profiles
or benthic flux measurements capture the immediate time scale of release, which may
not be directly coupled to organic oxidation. The 78-day residence time of water in
Tomales Bay provided a record of net reactions integrated through much of the
summer.
Si concentration decreased at the south of the bay (Fig. 4), while PO/- concentration did not (Fig. 5). This difference may reflect preferential settling of diatoms in deep
water immediately south of the sill; less dense, Si-free material may have drifted
further into the bay before settling. There must also have been local Si uptake in the
southernmost portion of the bay, by diatoms or localized emergent salt-marsh plants,
in order for the high Si concentration of stream water not to elevate the Si
concentration of the lower salinity water at station 20.
We infer the Si:P ratio of the material being regenerated from the Si:P export ratio
(2.4/0.12 = 20:1; Table 4). Grill and Richards (1964), Park (1967), Webb (1981),
Kaul and Froelich (1984), and Watson and Whitfield (1985) have derived Si:P ratios
of plankton, planktonic uptake, and planktonic regeneration between 20 and 24.
Brzezinski (1985) suggested that marine diatoms have a lower average Si content than
we have derived. Although he did not report Si:P ratios, the Si:N and Si:C ratios he
reported for diatoms grown in cultures are about half the value suggested by our data.
We acknowledge this discrepancy without having a ready explanation for it.
Decrease of DIN as water entered the bay could suggest N assimilation into organic
matter. This interpretation is not consistent with biogenic release of C, Si, and P, so we
look for an alternative. The most plausible alternative DIN sink is loss to gaseous N2 or
N20 by denitrification. Autotrophic uptake of DIN often favors NH4 + uptake over
NO]- (e.g. McCarthy, 1981; Wheeler, 1983). In fact, autotrophic NO]- assimilation
can be inhibited by the presence of NH4 + (e.g. Falkowski, 1983). The budgets (Table
4) indicate that net NO] - utilization preceded net NH4 + utilization. This pattern can
be interpreted to imply that NO] - may have been consumed by denitrification and
replenished by nitrification of NH4 + until the NH4 + was largely consumed. N02 -, an
intermediate product of both denitrification and nitrification, was present throughout
the bay, suggesting active transport through this reservoir.
Suspended material, sediments, and selected benthic plant samples were analyzed
for particulate C, N, and P (Table 5). The organic C:N:P molar ratio of suspended
material in the water column averaged 85:9:1, close to Ryther and Dunstan's (1971)
Smith et al.: Stoichiometry ofC, N, P & Sifluxes
1987]
445
Table 5. C, N, P composition of suspended material, sediments and benthic plants in Tomales
Bay. Suspended load data are in JLM; plant data are in mmol g-I; sediment data are in JLmol
g-I (mean ± std. dev.).
Material
Suspended load
(n = II)
Sediments
(n = 15)
Benthic plants
Zostera
(n = 4)
Viva
(n = 1)
Organic C
Total N
Total P
C:N:P
61
±
30
6.8
±
3.6
0.72
±
0.39
85:9:1
873
±
592
74
±
50
14
±
6
62:5:1
25.6
±
3.1
2.2
±
0.9
0.12
±
0.03
25.2
2.6
213:18:1
0.07
360:37:1
modified N:P ratio for coastal plankton. The organic C:N:P ratio of sediments was
62:5:1. The sediment C:P ratio is reasonably close to the ratio in suspended material,
but the sediments have a lower N content, relative to either P or C, than does the
suspended material. The C:N:P ratio of a small selection of benthic plants was about
240:20: 1, lower in N and especially P, relative to C, than suspended material. This
observation is consistent with the general composition of benthic plants (Atkinson and
Smith, 1983). We therefore assume that plankton-derived
material is the major
contributor to the sediment C:N:P ratio, with some differential N loss in the
sediments.
To test for internally consistent stoichiometry, we build a model with suspended
material being the fuel for net reactions. Table 6 summarizes the results of the inferred
total-system stoichiometry, scaled to C. The total dissolved C flux (.:lc,) is the sum of
Table 6. Partitioning total dissolved C, fixed N, P, and Si fluxes among component processes.
Rates in mmol m-2 d-1• Numerical values for calculations are in Table 4; see text for detailed
explanation.
Process
CaCO] precip. (.:l CC>
Oxic resp. (.:l C,)
Denitrification (.:l Cd)
CO2 gas flux (.:l CK)
Silica dissolution
Dissolved organic release
(.:lDGC)
Total non-conservative flux
(.:l C,)
How
estimated
Total
C
Fixed
N
Total
P
Total
Si
eq 9
eqs 12, 13
eqs 12, 13
eq 8
eq 6
-0.731
+ 7.668
+1.761
-10.534
0.000
0.000
+0.812
-1.222
0.000
0.000
0.000
+0.090
+0.021
.000
0.000
0.000
0.000
0.000
0.000
+2.439
eq 6
+0.590
+0.172
+0.009
0.000
eq 6
-1.246
-0.238
+0.120
+2.439
446
[45,2
Journal of Marine Research
"oxic respiration"
(AC,), DIC release by denitrification
(ACd), CaCOJ reactions
(6.CJ, gas flux (ACg) across the air-water interface, and DOC flux (ADOC):
(8)
Organic oxidation by sulfate reduction is not directly budgeted, although this
process is likely to be important in the sediments. H2S released to an oxygenated water
column would become oxidized back to SO/- (e.g. J~rgensen, 1977; Hansen et al ..
1978). "Oxic respiration" therefore includes both direct oxic respiration and sulfate
reduction resulting in H2S release to the water column. Any S bound into a mineral
sulfide will result in net TA elevation (e.g. Berner et al .• 1970). Mineral (e.g. Fe)
sulfide deposition would result in an underestimate of CaCOJ precipitation and organic
oxidation.
ACt is known directly (Table 4). 6.TA (also known) is a response to CaC03 reactions
and inorganic nitrogen flux (and, of course, sulfate reduction). N03 - and N02depress T A, while NH4 + raises it (Brewer and Goldman, 1976; Kinsey, 1978). Thus:
(9)
This equation is re-arranged and solved for ACe (Table 6). Any error in the total C
budget from mineral sulfide reactions should be small, because net TA flux is small.
The stoichiometric relationship among C, N, and P from oxic respiration and
denitrification can be described by the equations of Richards (1965), modified to
reflect the C:N:P ratio of suspended organic matter in Tomales Bay. Oxic respiration:
Denitrifica tion:
(CH20)gs(NHJ)9H3P04
+ 68HN03
=
85C02
+ 9NHJ +
H3P04
+
119 H20
+ 34N2
(11)
Oxidation of NH3 to NOJ - by nitrification is left out of these equations, because this
process results in no net gain or loss of DIN. It is instructive to show fixed N release
from organic matter to DIN during denitrification, rather than writing the reaction to
show the net change in fixed N.
It follows algebraically that inorganic P was released according to the P:C molar
ratio (PIC) of organic matter being oxidized by both oxic respiration and denitrification. There was also release of DOP from the organic matter:
(12)
Also by algebra, fixed N is released by oxic respiration and denitrification,
according to the N:C ratio (N IC) of the organic matter; N03 - in the water is
consumed in a 0.8:1 ratio with organic C oxidized by denitrification. The net change in
1987]
Smith et al.: Stoichiometry
of C, N, P & Si fluxes
dissolved fixed N (~N) includes N release from organic matter, NO) - utilization
denitrification, and direct release of DON:
447
by
(13)
Eqs. 12 and 13 can be solved simultaneously to yield estimates of ~Cr and ~Cd
(Table 6). Oxic respiration plus denitrification released about 9 mmol CO2 m-2d-l.
Net C release by denitrification totalled 1.8 mmol m-2d-1, only about 23% of the net
system carbon oxidation (Table 6). Fixed N consumed by denitrification totalled 1.4
mmol m-2d-l, and there was 0.2 mmol m-2d-1 of fixed N released by that denitrification. Thus, the net fixed N consumption by denitrification totalled 1.2 mmol m-2d-1
(Table 6), 150% of fixed N released during oxic respiration. The 50% excess dissolved
inorganic N required to support denitrification was imported from the ocean. Billen
(1978), Teal and Valiela (1978), Boynton et al. (1980), Nixon (1981), Hopkinson and
Wetzel (1982), Seitzinger et al. (1984), J~rgensen and S~rensen (1985), and
Hammond et al. (1985) have estimated estuarine and coastal denitrification rates
between 0.7 and 5 mmol m-2d-l. The calculated Tomales rate is towards the low end of
this range.
Eq. 8 is now rearranged and solved for ~Cg (Table 6). CO2 flux across the air-water
interface ( - 10.5 mmol m -2d -I) provides a check on the realism of the stoichiometric
derivation. Gas flux is a function of solubility (oc = 0.036 mmol m -3 /-Latm-I for CO2 at
the salinity and temperature of Tomales Bay), CO2 partial pressure difference between
the air and water (Pa - P", = 340 - 470 /-Latm), and the piston coefficient (k) (Smith,
1985):
(14)
By re-arrangement,
k = 2.2 m d -I. This value is higher than the value of 1 m d-I
derived from the empirical equation of Hartman and Hammond (1985) for a wind
speed of 4 m s -I, but within the scatter of the data which those authors summarize.
d. Sediment nutrient fluxes. Table 7 summarizes the sediment flux data. A few
obvious statistical outliers were removed, leaving 10-12 measurements in each sector.
Tabulated uncertainties
are based on variability among incubations, and do not
account for uncertainties associated with depth (probably primarily substratum).
Across the Inner Bay, the benthic fluxes of PO/- and Si were 2-5 times the
system-wide net fluxes (Table 4). The sediments showed release of NH4 +-a reversal
of the bay-wide net flux. T A flux was higher in the Inner Bay than the Outer Bay, we
suspect due to localized sulfate reduction (see section on stoichiometry). DIC release
was high, relative to the Inner Bay budget. DON flux was high, but variable; DOP flux
was low; DOC flux was not measured but is assumed on the basis of bay-wide net
DOC:DON:DOP
flux ratios to be small relative to DIC flux. These results are
448
[45,2
Journal of Marine Research
Table 7. Nutrient fluxes from sediments, as measured with chamber incubations. Fluxes in
mmol m-2d-1 (meq m-2d-1 for TA). Mean ± std. dey. (n).
Material
(Station)
Outer
(3,5)
Central
(10, 10)
South
(16, 18)
Inner-bay
mean*
Si
5.94 ± 5.10
(12)
5.75 ± 2.20
(11)
1.16 ± 5.28
(12)
4.73 ± 3.16
PO/-
0.22±0.15
(10)
0.15 ± 0.27
(12)
0.37 ± 0.31
0.29 ± 0.21
(10)
-0.09 ± 0.09
(12)
0.20 ± 0.23
-0.18 ± 0.26
(12)
0.05 ± 0.08
(12)
-0.13 ± 0.27
0.19 ± 0.22
-0.10 ± 0.20
(11)
0.02 ± 0.05
(12)
2.12 ± 1.70
(12)
0.06 ± 1.79
(11)
2.10 ± 2.48
0.03 ± 0.07
(9)
0.00 ± 0.06
(12)
2.17 ± 1.01
(10)
3.25 ± 3.23
(12)
5.45 ± 3.39
0.06 ± 0.08
(12)
0.03 ± 0.05
(12)
1.64 ± 0.91
(11)
2.07 ± 1.21
( 12)
3.80 ± 1.52
0.04 ± 0.07
0.7 ± 9.1
(6)
28.2 ± 28.6
(6)
0.0
28 ± 29
14.3 ± 12.1
(9)
35.6 ± 24.6
(12)
0.0
36 ± 25
14.2 ± 5.6
(10)
53.9 ± 11.0
(10)
0.0
54 ± 11
14.3 ± 11.0
DOP
tot. P
NOJN02NH4+
DON
tot. N
TA
DIC
DOC**
tot. C
-0.06 ± 0.09
0.13 ± 0.24
0.01 ± 0.06
2.05
±
0.99
2.99 ± 2.91
5.09 ± 3.08
39.7 ± 22.3
0.0
40 ± 23
*Area-weighted means; data from Table 1.
**Assumed = 0; not measured.
consistent with other studies of benthic flux (e.g. Elderfield et al., 1981; Callender and
Hammond, 1982; Hopkinson and Wetzel, 1982; Hammond et al., 1985).
A particularly interesting aspect of the chamber flux data is that the north sector
(Outer Bay, Table 1), and the central sector resemble one another more so than do the
central and southern sectors: the major difference (to the water chemistry) between the
Inner and Outer Bay is apparently one of water exchange, not rates of benthic
regeneration. Si and PO/- fluxes were rapid in the two northern sectors and then
decreased to the south, introducing large standard deviations into the calculated Inner
Bay rates. Mean PO/- flux in the southern sector was negative. Rapid Si and PO/release in the north, decreasing to the south, is consistent with the general shapes of the
bay-wide Si and PO/- concentration gradients (Figs. 4, 5).
Because the benthic chambers were opaque, photo-autotrophic processes are not
likely to account for the lower Si release and reversal of PO/- flux in the southern
Smith et al.: Stoichiometry of C. N. P & Si fluxes
1987]
449
Table 8. Estimates of water-column net primary production based on 14C uptake experiments.
Station productivity has been weighted by sector mean depth and light distribution through
the water column. Estimates of nutrient uptake are scaled by the estimated net C:N:P:Si flux
ratio. Also included are estimates of productivity in the Inner Bay in samples receiving NH4 +
spike.
2
Station
8
(Outer)
Photic depth (m)
6.9
Sector mean depth (m)
2.2
Sector net production (station productivity
distribution through the water column)
(mmol C m-2d-l)
23.6
(mmol N m-2d-1)
2.50
(mmol P m-2d-1)
0.28
(mmol Si m-2d-l)
5.55
Station net productivity
(mmol C m-2d-l)
*Weighted
with NH4
+
(Central)
16
(South)
5.0
5.2
4.4
1.0
Inner bay
mean*
weighted by sector mean depth and light
53.8
5.70
0.63
12.66
11.8
1.25
0.14
2.78
44.5
4.71
0.52
10.46
spike
178
62
by sector area, Table 1.
sector. The lower Si flux to the south might result from the previously suggested
preferential sedimentation of diatom frustules as water crosses the sill, into the Inner
Bay. We suspect that the reversal of the P flux might represent local inorganic P
adsorption or precipitation (as discussed above), perhaps because of the Fe content and
sorptive capacity of seasonally deposited terrigenous sediments. This phenomenon
appears to be of limited significance to the bay-wide budget because of the relatively
small area of the southern sector (Table 1).
We also calculated diffusion-supported fluxes for four cores, using nutrient concentration gradients in the upper 2 cm of the cores, average measured porosity (70%), and
molecular diffusivity (from Li and Gregory, 1974, for Si, H2P04 -, and NH4 +; and
Lerman, 1979, for NO) -). Mean diffusion-supported fluxes (in mmol m-2d -I) were:
Si = 0.18; PO/- = 0.06; NH/ ~ 0.31; NO)- = 0.004. These diffusion- supported
fluxes are low (by 3-30 times) in comparison to measured fluxes (Table 7). Such
discrepancies have been repeatedly noted in other systems; we are inclined to favor the
view of Kemp et al. (1982), that much of the benthic nutrient release is from a surface
flocculant layer, rather than from the sediment column.
e. Water-column primary production. Table 8 summarizes phytoplankton net primary production. The table indicates the estimated photic-zone depth (consistently
greater than sector mean depth) and integrated net production over mean water depth
in each sector. Inner-bay primary production can be used to estimate the incorporation
of N, P, and Si into phytoplankton biomass. The estimates of nutrient uptake by
[45,2
Journal of Marine Research
450
Table 9. Summary of water-column PO.3- cycling measurements. Details of uptake and
regeneration calculations are explained in text.
PO/Station
~M*
~M/hr**
Uptake
~M/hr
Regeneration
~M/hr
7 Oct
2
12
16
0.57
3.81
4.56
n.s.***
n.s.
n.s.
0.013
0.012
0.010
n.s.
n.s.
n.s.
8 Oct
2
12
18
0.89
4.26
4.60
n.s.
n.s.
n.s.
0.011
0.010
0.018t
n.s.
n.s.
n.s.
9 Oct
0
12
16
1.89
4.38
4.93
n.s.
n.s.
n.s.
0.005
0.015t
0.012t
n.s.
n.s.
n.s.
Date
Mean
±
std. dey
Outer
Central
South
0.010
0.012
0.013
±
±
±
0.004
0.003
0.004
0.010:j:
0.012+
0.013:j:
*Initial concentration.
**Regression slope.
***n.s. ~ not statisticaIly significant (P = 0.05).
tBased on 32p_PO/- uptake from first hour only. Time trend becomes non-linear thereafter.
:j:Assumednot different from uptake, as discussed in text.
phytoplankton are based on the C:N:P:Si flux ratio which we have previously
established (85:9: 1:20). If Brzezinski's (1985) Si data for diatoms were used instead of
our flux-derived estimate, phytoplankton uptake of Si would be reduced by approximately one half. The point to note is that estimated phytoplankton uptake rates are all
higher than the absolute baywide net fluxes (Table 4), but of similar magnitude to the
benthic fluxes (Table 7).
In addition to the incubations of unmodified water samples, samples at the two Inner
Bay sites (where DIN was low) were spiked to approximately 20 J.LM NH/, prior to
incubation. There was a substantial increase in productivity in bottles spiked with
NH. +, suggesting that phytoplankton primary production at the time of these
incubations was limited by the availability of DIN.
f Water-column Nand P cycling. Table 9 summarizes data from the water-column
PO/- cycling experiments. There was insignificant net change in PO/- concentrations over the 4-hour incubations, so gross PO/- regeneration rates cannot be
distinguished from 0 (Harrison, 1983; Eq. 1). Because there is no significant net
change in PO/-, gross regeneration is also indistinguishable from uptake. Gross
uptake of PO/- can be measured even though regeneration cannot, because uptake is
451
Smith et al.: Stoichiometry ofC, N, P & Sijluxes
1987]
Table 10. Summary of water-column NH4 + cycling measurements. Details of uptake and
regeneration calculations are explained in text.
Station
~M**
NH4+
~M/hr***
Uptake
~M/hr
Regeneration *
~M/hr
7 Oct
2
12
16
2.11
0.90
2.81
-0.105
-0.165t
-0.1 lOt
0.142
0.301
0.110
0.037
0.136
n.s.
8 Oct
2
12
18
3.44
1.16
1.16
-0.086
n.s.:!:
-0.165
0.301
0.132
0.165
0.215
0.063
n.s.
9 Oct
0
12
16
4.06
5.68
2.90
n.s.
0.083
0.003
n.s.
n.s.
n.s.
n.s.
0.083
0.003
Date
Mean
±
std. dey
Outer
Central
South
0.148
0.144
0.110
±
±
±
0.151
0.151
0.055
0.084
0.094
0.001
±
±
±
0.115
0.Q38
0.002
*Using Blackburn (1979) if significant change in NH4 +; otherwise Glibert et al. (1982).
**Initial concentration.
***Regression slope.
to-2 hr data only. Time trend non-linear thereafter.
:!:n.s.= not statistically significant (P ~ 0.05); use value of 0 for calculating means.
based on incorporation of 32p into particulate
matter (Harrison,
1983; Eq. 2),
analytically a sensitive measurement.
Both uptake and regeneration of NH4 + are individually calculable (Table 10).
NH4 + uptake clearly exceeded regeneration in the south sector and showed such a
tendency elsewhere. NH4 + uptake and regeneration data show considerable variability, both among sites and among days. We suspect that the variability primarily reflects
variable amounts of wind-stirred resuspended material in the water column.
Several conclusions about water-column nutrient turnover can be derived when
24-hour uptake rates for NH4 + and PO/- are averaged over the Inner Bay (Table 11).
The NH4+ :PO/- uptake ratio is about 12:1, in reasonable agreement with the 9:1
Table II. Estimates of daily water-column POl- and NH4 + uptake (based on assumption of
24-hour uptake). Uptake data from Tables II and 12 and hypsographic data from Table 1.
Sector
Outer
Central
South
Inner-Bay mean
mmol m-J d-1
NH4+
POl0.24
0.29
0.31
3.55
3.46
2.64
pol0.53
1.28
0.31
1.22
mmol m-2d-1
NH4+
7.8
15.2
2.6
14.4
N:P
15
12
8
12
452
Journal of Marine Research
[45,2
Table 12. Potential denitrification and relative nitrogen fixation in various materials.
Location
Sta.5
Material
sediment
sediment
Sta.8
Sta. 3
Sta.6
Sta. 10
Sta. 18
sediment
sediment
grass
sediment
grass/algae
sediment
grass
sediment
Process
Denit.
(NO)-loss)
(90 min)
Denit.
(NO)- loss)
(200 min)
Denit.
(N20 uptake)
(16 hr)
N2
N2
N2
N2
N2
N2
N2
fix.
fix.
fix.
fix.
fix.
fix.
fix.
Rate
mmol m-2d-1
37
47
48
0
0
0
0
low
low
low
suspended N:P ratio in the bay (Table 5). When NH4 + and PO/- uptake rates are
compared with the nutrients inferred to be required for primary production (Table 8),
the Nand P fluxes are 2-3 times as high as predicted from primary production.
We assume that this difference in part reflects the difference between gross uptake
(for Nand P) and net uptake (for C). A second consideration is the role of heterotroph
uptake in accounting for the observed uptake. Inventories of fluxes to and from all of
the Nand P reservoirs would be required to address this point. The independently
performed primary production and nutrient turnover experiments may also individually contain artifacts which preclude close comparison. The important point to note is
that the water-column turnover rates are substantially more rapid than the system net
rates (Table 4).
g. Biochemical nitrogen transformation reactions. An examination of N2 fixation and
denitrification was conducted in order to establish whether these processes could be
demonstrated in zones consistent with the system-wide analyses.
Denitrification potential, shown in Table 12, was high at both station 5, which was in
the region of high NO) - concentrations (Fig. 6) and station 8, in the region of low
N03 - concentrations. As demonstrated by a NO) - uptake experiment at station 5,
rates obtained over 90 minutes were about as fast as rates over 200 minutes. This
observation suggests that there was little or no initial lag period of slow denitrification
once N03 - was added to the incubation chambers. Results of a 16-hour N20 uptake
experiment for sediments at station 8 yielded about the same rates as those obtained at
1987]
Smith et al.: Stoichiometry ofC. N, P & Sij/uxes
453
station 5, suggesting that the microbial community is capable of high rates of
denitrification throughout much of Tomales Bay. It should be emphasized that the
rates obtained are potentials only, and should not be construed to represent in situ
rates.
Sediment N2 fixation was assayed at four stations. One Outer Bay station (3) and
one transition station (6), where DIN concentrations were fairly high but variable,
showed negligible N2 fixation. Two Inner Bay stations (10,18), where DIN concentrations were generally low, showed low but measureable N2 fixation. Pieces of drift
seagrass and algae showed much the same pattern. N2 fixation seems likely to occur at
low but significant rates throughout much of the Inner Bay.
We are aware of the discussion surrounding the acetylene blockage technique and
NO] - uptake as measures of denitrification (e.g. Seitzinger et al., 1984) and with
concerns over the stoichiometry for conversion from acetylene reduction to N2 fixation
(e.g. Capone, 1983). We emphasize that we have used these field assays to establish a
preliminary understanding of the potential for these N transformation reactions in
Tomales Bay, and some idea as to the distribution of that potential.
6. Discussion
This study was designed to test the generality of Smith's (1984) hypothesis that P,
not N, limits organic carbon production in confined aquatic systems. It was assumed
that Tomales Bay would be net autotrophic, hence to have a net demand for Nand P to
support the production of new organic carbon. That assumption proved to be incorrect
at the time of this investigation, so questions about the supply of nutrients to support
net ecosystem production remain unaddressed by the investigation. Hypotheses to
explain inorganic nutrient limitation of net ecosystem organic production cannot be
addressed in a system which consumes, rather than produces, organic carbon.
We approached the question of net ecosystem metabolism from both a total-system
mass-balance analysis and analyses of major component fluxes. Figure 8 is a summary
of the total-system budgets and major component fluxes. While the component-flux
studies were not as comprehensive as the analysis of total-system fluxes, these
component studies were sufficient to make the following point. There is no problem in
accounting for net ecosystem fluxes by the rates of nutrient turnover through
component cycles within the system.
Our data address several scales of turnover cycles embedded within the net
ecosystem fluxes. Water-column nutrient turnover (Table 11) was more rapid than net
nutrient uptake by phytoplankton production (Table 8). The cycle between net
water-column uptake (Table 8) and net sediment release (Table 7) was rapid in
comparison with net nutrient exchange between the Inner Bay and adjacent reservoirs
(Tables 4, 6). Transformations between fixed and gaseous N were not quantified
(Table 12), but were apparently capable of rates which were rapid in comparison with
net exchange (Table 6). It is statistically difficult to evaluate slow net system
454
Journal of Marine Research
[45,2
Figure 8. Simplified summary diagrams of C, N, P, and Si fluxes in Inner Tomales Bay. The
cross-hatched column represents the component flux calculations, with the inner box being the
water-column fluxes. Solid arrows are dissolved-load fluxes; dashed arrows are particle fluxes;
and dotted arrows are gas fluxes. Arrows not in contact with the hatched area represent
total-system budgetary calculations; arrows to, from, or within the hatched areas are based on
the component budgets. Note that the component fluxes do not "add up" to the net system
fluxes or necessarily reflect their direction; nevertheless the component fluxes are of sufficient
magnitude to account for those net fluxes.
performance as the sum of rapid, cyclical fluxes. Had we only approached this system
component by component, it would have been difficult to evaluate rates, or even
directions, of net fluxes for the entire ecosystem.
The budgetary analyses integrate temporal and spatial characteristics of the system.
We conclude with some confidence that Inner Tomales Bay was exporting dissolved P
and Si, exporting gaseous CO2, and importing dissolved fixed N at the time of our
study. Regeneration of dissolved nutrients from organic matter, loss of fixed N to
gaseous N, and gas evasion are consistent with these fluxes. This net system
performance is neither apparent from nor inconsistent with the component studies.
Reservoirs of particulate organic matter outside Tomales Bay and hydrographic
characteristics of the Inner Bay as a settling basin give the bay potential to trap organic
matter, become heterotrophic, and release nutrients. Without data over an annual
cycle, we cannot be sure whether the particle supply supporting the calculated
1987]
Smith et at.: Stoichiometry ofe, N, P & Sijluxes
455
heterotrophy represents ongoing (tidal) particle delivery; a seasonal sequence of net
production and sedimentation, followed by regeneration; or regeneration from previously sedimented terrigenous input. However our "snapshot in time" did have an
"exposure time" of about 80 days (i.e. about 20% of the year; 40% of the dry, upwelling
season). Thus the heterotrophy describes a substantial fraction of an annual cycle.
It would be inappropriate to apply the box model used here to estimate particle
fluxes. Longitudinal gradients of particle concentration are not reliable indicators of
net fluxes for materials which are readily sedimented and resuspended in response to
spatially and temporally varying turbulence. The low mixing coefficient and deep
basin of Inner Tomales Bay are qualitative clues that dePosition might be expected
there; but extensive measurements of particle transport, sedimentation, and resuspension would be required to quantify this deposition. Welsh et at. (I982) observed: "The
prospects for securing estimates (of net particulate transport) with any degree of
precision appear to be dim, given the methodological problems."
Hopkinson (1985) recognized the difficulties of inferring trophic status from
detrital flux. He advocated the "recipient system" approach: "By measuring the level
at which community respiration exceeds primary production, the import of allochthonous material is directly observed." Hydrographic budgets for dissolved materials are
recipient-system analyses of the entire system at once. The budgets yield unambiguous
statements about net fluxes of materials; simultaneous stoichiometric analyses ofC, N,
P, and Si provide more insight into biogeochemical processes than would separate
budgets.
An estimate of net organic C metabolism for the ecosystem was the "desired
product" of the hydrographic budgets, but DIC changes include CaC03 reactions and
gas flux as well as C metabolism. CaC03 reactions can often be unambiguously
deciphered by changes in TA (e.g. Smith and Jokiel, 1978; Smith and Atkinson, 1983).
In this system, CaC03 reactions were not unambiguously separable from mineral
sulfide deposition, but CaC03 precipitation seemed to be minor. CO2 gas flux can be a
quantitatively important term in a total C budget, and this flux cannot be directly
Qerived from hydrographic analysis of C flux alone. In order to constrain organic C
metabolism, either CO2 gas flux must be estimated directly (e.g. Hartman and
Hammond, 1985) or organic metabolism must be constrained by stoichiometric
considerations (this study; Nixon and Pilson, 1984).
A dissolved P budget scaled by the organic C:P ratio is probably the simplest single
measure of ecosystem net metabolism, an idea which has been aired previously (e.g.
Postma, 1954; Smith and Jokiel, 1978; Atkinson, 1981). Elevated summertime P
concentrations in other estuaries (Taft and Taylor, 1976) may also represent local
heterotrophy on some time scale, although abiotic P reactions cannot be ruled out by a
P budget alone.
Dissolved Si budgets can also be useful in deciphering biogeochemical fluxes, despite
456
Journal of Marine Research
[45,2
the potential importance of mineral-water reactions. Concordance between biochemically expected (Grill and Richards, 1964; Webb, 1981) and environmentally calculated (Park, 1967; Kaul and Froelich, 1984; Watson and Whitfield, 1985; this study)
Si:P ratios is circumstantial evidence that biochemical processes dominate both Si and
P fluxes.
Hydrographic
budgets for fixed N alone contain ambiguous information about
system-level net organic C metabolism. Some autotrophic ecosystems largely meet
their demand for N to assimilate into net organic production by fixation of N 2 gas (e.g.
Smith, 1984). Heterotrophic systems can also have their N budgets dominated by gas
transfer. Even the modest rate of organic C oxidation by denitrification for Inner
Tomales Bay apparently converted available dissolved and particulate fixed N to
gaseous N. Sulfate reduction probably occurs in most shallow-water marine environments, yet denitrification is energetically favored over sulfate reduction (e.g. Reeburgh, 1983). Such ecosystems should therefore show a tendency towards heterotrophic consumption of available fixed N.
Denitrification may be limited by availability of either organic substrate (Kaspar,
1982) or NO) - (Oremland et al., 1984). Altering the supply of either organic substrate
or dissolved fixed N by changing the hydrographic regime can therefore alter the
denitrification
rate. However, it is difficult to imagine that greatly altering the
system-wide denitrification rate (a relatively slow rate of fixed N consumption) would
have much effect on fixed N utilization by primary production (a much more rapid
process relying largely on recycled nutrients).
Denitrification
and limitation of
primary production by dissolved fixed N in aquatic ecosystems may therefore be
symptoms of other system-scale constraints
on net C metabolism, rather than
themselves being system-level controls of net C metabolism.
Acknowledgments. We thank S. Obrebski (TCES) for logistic assistance and discussions
about the natural history of Tomales Bay; T. Walsh (UH), N. Dudek (OSU), and S. Vink (UH)
for technical support; R. Casassa for collecting weather data; P. Tyra and J. Null, (National
Weather Service) for assistance with weather data; K. Markham (USGS) for stream data; and
the following individuals for comments on aspects of this study: M. Atkinson, T. Berman,
K. Chave, W. Green, Y.-H. Li, D. Karl, and F. Mackenzie. The Tiburon Center for
Environmental Studies (TCES) provided a boat and laboratory facilities. The research was
sponsored by NSF Grant Nos. OCE 85-00679 (awarded to SVS) and OCE 82-14837 (awarded
to JTH). Contribution No. 732 of the Hawaii Institute of Marine Biology, and 1849 of the
Hawaii Institute of Geophysics.
Abbott, M. R. and P.
an upwelling event.
Armstrong, F. A., P.
matter in seawater
481-483.
REFERENCES
M. Zion. 1985. Satellite observations of phytoplankton variability during
Cont. Shelf Res., 4, 661-680.
M. Williams and J. D. H. Strickland. 1966. Photo-oxidation of organic
by ultraviolet radiation, analytical and other applications. Nat., 211.
1987]
Smith et at.: Stoichiometry ofe, N, P
&
Sifiuxes
457
Atkinson, M. J. 1981. Phosphorus metabolism of coral reef flats. Ph.D. thesis, U. Hawaii,
1-90.
Atkinson, M. J. and S. V. Smith, 1983. C:N:P ratios of benthic marine plants. Limnol.
Oceanogr.,26, 1074-1083.
Bakun, A. 1975. Daily and weekly coastal upwelling indices, west coast of North America,
1946-71. NOAA Technical Report NMFS SSRF-693, 114 pp.
Berner, R. A., M. R. Scott and C. Thomlinson. 1970. Carbonate alkalinity in the pore waters of
anoxic marine sediments. Limnol. Oceanogr., 15, 544-549.
Billen, G. 1978. A budget of nitrogen recycling in North Sea sediments off the Belgian coast.
Est. Coast. Mar. Sci., 7, 127-146.
Blackburn, T. H. 1979. Method for measuring rates of NH4 + turnover in anoxic marine
sediments, using a 14N-NH4+ dilution technique. Appl. Environ. Microbiol., 37, 760-765.
Boyle, E. A., R. Collier, A. T. Dengler, J. M. Edmond, A. C. Ng and R. F. Stallard. 1974. On the
chemical mass-balance in estuaries. Geochim. Cosmochim. Acta, 38. 1719-1728.
Boynton, W. R., W. M. Kemp and C. Osborne. 1980. Nutrient fluxes across the sediment-water
interface in the turbid zone of a coastal plain estuary, in Estuarine Perspectives, V. S. Kennedy, ed., Academic, NY, 93-109.
Brewer, P. G. and J. C. Goldman. 1976. Alkalinity changes generated by phytoplankton growth.
Limnol. Oceanogr., 21, 108-117.
Brzezinski, M. A. 1985. The Si:C:N ratio of marine diatoms: interspecific variability and the
effect of some environmental variables. J. Phycol., 21, 347-357.
Callender, E. and D. F. Hammond. 1982. Nutrient exchange across the sediment-water
interface in the Potomac River estuary. Est. Coast. Shelf Sci., 15, 395-413.
Capone, D. G. 1983. Benthic nitrogen fixation, in Nitrogen in the Marine Environment,
E. J. Carpenter and D. G. Capone, eds., Academic, NY, 105-137.
Cole, B. E. and J. E. Cloern. 1984. Significance of biomass and light availability to phytoplankton productivity in San Francisco Bay. Mar. Ecol.-Prog. Ser., 17, 15-24.
Daetwyler, C. C. 1966. Marine geology of Tomales Bay, California. Pacific Marine Station
Research Report. No.6., 169 pp.
Dudek, N., M. A. Brzezinski and P. A. Wheeler. 1986. Recovery of ammonium nitrogen by
solvent extraction for the determination of relative lSN abundance in regeneration experiments. Mar. Chern., 18, 59-69.
Elderfield, H., N. Luedtke, R. J. McCaffrey and M. Bender. 1981. Benthic flux studies in
Narragansett Bay. Am. J. Sci. 281,768-787.
Falkowski, P. G. 1983. Enzymology of nitrogen assimilation, in Nitrogen in the Marine
Environment, E. J. Carpenter and D. G. Capone, oos., Academic, NY, 731-762.
Fehler, S. W. 1981. Nutrient chemistry, in A Climatology and Oceanographic Analysis of the
California Pacific Outer Continental Shelf Region, F. A. Godshall and R. G. Williams, oos.,
U.S. Dept. Commerce, p. 6-1 to 6-33.
Glibert, P. M., F. Lipschultz, J. J. McCarthy and M. A. Altabet. 1982. Isotope dilution models
of uptake and remineralization of ammonium by marine plankton. Limnol. Oceanogr., 27,
639-650.
Grill, E. V. and F. A. Richards. 1964. Nutrient regeneration from phytoplankton decomposition
in seawater. J. Mar. Res., 22. 51-69.
Hammond, D. E. and others. 1985. Benthic fluxes in San Francisco Bay. Hydrobiol., 129,
69-90.
Hansen, M. H., K. Ingvorsen and B. B. J~rgensen. 1978. Mechanisms of hydrogen sulfide
release from coastal marine sediments to the atmosphere. Limnol. Oceanogr., 23, 68-76.
458
Journal of Marine Research
[45,2
Harrison, W. G. 1983. Uptake and recycling of soluble reactive phosphorus by marine
microplankton. Mar. Ecol.-Prog. Ser., 10. 127-135.
Hartman, B. and D. E. Hammond. 1985. Gas exchange in San Francisco Bay. Hydrobiol., J 29,
59-68.
Hopkinson, C. S. Jr. 1985. Shallow-water benthic and pelagic metabolism: evidence of
heterotrophy in the nearshore Georgia Bight. Mar. BioI., 87, 19-32.
Hopkinson, C. S. Jr. and R. L. Wetzel. 1982 In situ measurements of nutrient and oxygen fluxes
in a coastal marine benthic community. Mar. Eco1.-Prog. Ser., 10, 29-35.
J~rgensen, B. B. 1977. The sulfur cycle of a coastal marine sediment (Limfjorden, Denmark).
Limno1. Oceanogr., 22,814-832.
J~rgensen, B. B. and J. S~rensen. 1985. Seasonal cycles of O2, NO) -, and SO/- reduction
maximum in spring. Mar. Eco1.-Prog. Ser., 24, 65-74.
Kaspar, H. F. 1982. Denitrification in marine sediment: measurement of capacity and estimate
of in situ rate. Appl. Environ. Microbiol., 22, 43-51.
Kaul, L. W. and P. N. Froelich, Jr. 1984. Modeling estuarine nutrient geochemistry in a simple
system. Geochim. Cosmochim Acta, 48. 1417-1433.
Kemp, W. M., R. L. Wetzel, W. R. Boynton, C. F. D'Elia and J. C. Stevenson. 1982. Nitrogen
cycling and estuarine interfaces: some current concepts and research directions, in Estuarine
Comparisons, V. S. Kennedy, ed., Academic, NY, 209-230.
Kinsey, D. W. 1978. Alkalinity changes and coral reef calcification. Limnol. Oceanogr., 23.
989-991.
Klump, J. V. and C. S. Martens. 1981. Biogeochemical cycling in an organic rich coastal marine
basin-II. Nutrient sediment-water exchange processes. Geochim. Cosmochim. Acta, 45.
101-121.
Krom, M. D. and R. A. Berner. 1981. The diagenesis of phosphorus in a nearshore sediment.
Geochim. Cosmochim. Acta, 45, 207-216.
Lerman, A. 1979. Geochemical processes: water and sediment environments. WileyInterscience, NY, 481 pp.
Li, Y.-H. and S. Gregory. 1974. Diffusion of ions in sea water and in deep-sea sediments.
Geochim. Cosmochim. Acta, 38. 703-714.
Loder, T. C. and R. P. Reichard. 1980. The dynamics of conservative mixing in estuaries.
Estuaries, 4, 64-69.
McCarthy, J. J. 1981. Uptake of major nutrients by estuarine plants, in Estuaries and Nutrients,
B. J. Neilson and L. E. Cronin, eds., Humana, Clifton, NJ, 139-163.
Martens, C. S., R. A. Berner and J. K. Rosenfeld. 1978. Interstitial water chemistry of anoxic
Long Island Sound sediments. 2. Nutrient regeneration and phosphate removal. Limnol.
Oceanogr., 23, 605-617.
Nixon, S. W. 1981. Remineralization and nutrient cycling in costal marine ecosystems, in
Estuaries and Nutrients, B. J. Neilson and L. E. Cronin, eds., Humana, Clifton, NJ,
111-138.
Nixon, S. W. and M. E. Q. Pilson. 1984. Estuarine total system metabolism and organic
exchange calculated from nutrient ratios: an example from Narragansett Bay, in The Estuary
as a Filter, V. S. Kennedy, ed., Academic, Orlando, FL., 261-290.
Officer, C. B. 1979. Discussion of the behavior of nonconservative dissolved constituents in
estuaries. Est. Coast Mar. Sci., 9. 91-94.
Officer, C. B. and D. R. Lynch. 1981. Dynamics of mixing in estuaries. Est. Coast Shelf Sci., 12,
525-533.
Oremland, R. S., C. Umberger, C. H. Culbertson and R. L. Smith. 1984. Denitrification in San
Francisco Bay intertidal sediments. Appl. Environ. Microbiol., 47, 1106-1112.
1987]
Smith et al.: Stoichiometry
ofe, N, P & Sifluxes
459
Park, K. 1967. Nutrient regeneration and preformed nutrients off Oregon. Limnol. Oceanogr.,
12, 353-356.
Pilson, M. E. Q. 1985. On the residence time of water in Narragansett Bay. Estuaries, 8,2-14.
Postma, H. 1954. Hydrography of the Dutch Wadden Sea. Arch. Neerlandaises Zool., /0,
405-511.
Reeburgh, W. S. 1983. Rates of biogeochemical processes in anoxic sediments. Ann. Rev. Earth
Planet. Sci., II. 269-298.
Richards, F. A. 1965. Anoxic basins and fjords, in Chemical Oceanography, v.l, 1st ed.,
J. P. Riley and G. Skirrow, eds., Academic, NY, 611-645.
Ryther, J. H. and W. M. Dunstan. 1971. Nitrogen, phosphorus, and eutrophication in the
coastal marine environment. Sci., 171, 1008-1013.
Seitzinger, S. P., S. W. Nixon and M. E. Q. Pilson. 1984. Denitrification and nitrous oxide
production in a coastal marine ecosystem. Limnol. Oceanogr., 29, 73-83.
Skirrow, G. 1975. The dissolved gases-carbon dioxide, in Chemical Oceanography, v.2, 2nd ed.,
J. P. Riley and G. Skirrow, eds., Academic, NY, 1-92.
Smith, E. H., R. G. Johnson and S. Obrebski. 1971. Physical, chemical, microbiological and
hydrographic characteristics of Tomales Bay. Pacific Marine Station Research Report. No.9,
70 pp + 4 separately numbered appendices.
Smith, S. V. 1984. Phosphorus versus nitrogen limitation in the marine environment. Limnol.
Oceanogr., 29, 1149-1160.
-1985. Physical, chemical and biological characteristics of CO2 gas flux across the air-water
interface. Plant, Cell Environ., 8, 387-398.
Smith, S. V. and M. J. Atkinson. 1983. Mass balance of carbon and phosphorus in Shark Bay,
Western Australia. Limno\. Oceanogr., 28, 625-639.
Smith, S. V. and P. L. Jokie\. 1978. Water composition and biogeochemical gradients in the
Canton Atoll lagoon. Atoll Res. Bull., 221, 15-53.
Smith, S. V. and D. W. Kinsey. 1978. Calcification and organic carbon metabolism indicated by
carbon dioxide. Monogr. Oceanogr. Methodol., 5, 469-484.
Sverdrup, H. U., M. W. Johnson and R. H. Fleming. 1942. The Oceans: their Physics,
Chemistry and General Biology, Prentice-Hall, Englewood Cliffs, NY, 1-1087.
Strickland, J. D. H. and T. R. Parsons. 1972. A practical handbook of seawater analysis. Bull.
Fish. Res. Bd. Can., 167, 1-311.
Taft, J. L. and W. R. Taylor. 1976. Phosphorus dynamics in some coastal plain estuaries, in
Estuarine Processes, v. 1., M. Wiley, ed., Academic, NY, 79-89.
Teal, J. M. and I. Valiela. 1978. Nitrogen budget of a coastal marine ecosystem. Rapp. P.-v.
Reun. Cons. int. Explor. Mer, 173, 101-105.
Watanabe, Y. and S. Tsunogai. 1984. Adsorption-desorption control of phosphate in anoxic
sediments of a coastal sea, Funka Bay, Japan. Mar. Chern., 15, 71-83.
Watson, A. J. and M. Whitfield. 1985. Composition of particles in the global ocean. Deep-Sea
Res.,32,1023-1039.
Webb, K. L. 1981. Conceptual models and processes of nutrient cycling in estuaries, in Estuaries
and Nutrients, B. J. Neilson and L. E. Cronin, eds., Humana, Clifton, NJ, 25-46.
Webb, K. L., W. D. Dupaul, W. Wiebe, W. Sottile and R. E. Johannes. 1975. Enewetak
(Eniwetok) Atoll: aspects of the nitrogen cycle on a coral reef. Limnol. Oceanogr., 20,
198-210.
Weiss, R. F. 1974. Carbon dioxide in water and seawater: the solubility of a non-ideal gas. Mar.
Chern., 2,202-215.
Welsh, B. L., R. B. Whitlatch and W. F. Bohlen. 1982. Relationship between physical
460
Journal of Marine Research
[45,2
characteristics and organic carbon sources as a basis for comparing estuaries in Southern New
England, in Estuarine Comparisons, V. S. Kennedy, ed., Plenum, NY, 53-67.
Wheeler, P. A. 1983. Phytoplankton nitrogen metabolism, in Nitrogen in the Marine Environment, E. J. Carpenter and D. G. Capone, eds., Academic, NY, 309-346.
Wollast, R. 1978. Modelling of biological and chemical processes in the ScheIdt Estuary, in
Hydrodynamics of Estuaries and Fjords, J. C. J. Nihoul, ed., Elsevier, Amsterdam, 63-77.
Received: 28 May. 1986; revised: 23 March,1987.